SPECTRAL
ENERGY DISTRIBUTION ACETYLENE FLAME.*
IN THE
BY
W. Associate
W.
COBLENTZ,
Physicist,
Bureau
Ph.D. of Standards
IN previous papers the writer l published data on the spectral energy distribution of the acetylene flame. In one of these papers 2 an estimate is given of the accuracy It is based upon the attainable in making the observations. deviations of the observations from the mean value; and it is shown that whereas in the red end of the spectrum an accuracy of 0.5 per cent. is attainable (provided the flame stays constant) that an accuracy of IO per cent. is hardly attainable in the violet. Unfortunately no mention was made of the difficulties and uncertainties in reducing these observed data from the prismatic into The purity and slit width factors were the normal spectrum. probably accurate to I per cent. The greatest uncertainty was the correction for absorption in the glass, prism or the silver mirrors (or both) in the violet end of the spectrum. Corrections must be made also for diffuse light. To overcome the difficulties of absorption a spectrometer, consisting of a quartz prism and two plano-convex lenses, was used for measurement in the violet. Not being achromatic, a correction had to be made for change in aperture, with change in focal length for different parts of the spectrum. A recent examination shows that the published data were probably overcorrected (in the violet) for change in aperture; and the utility of such an instrument for making such measurements is to be questioned. The spectral energy data published by the writer 3 were obtained from a composite curve consisting of measurements from 0.5~ to 75~ made (first) with a mirror spectrometer and fluorite prism, and (second) a glass lens spectrometer with flint glass prism (the absorption of which has been determined), supplemented by measurements, from 0.4~ to 0.5~~ made with (third) the quartz lens spectrometer just mentioned. * Communicated
by the
Author.
’Coblentz,
Bul. Bur. Stds., 7. p. 253, 1911 ; 13, p. 355, 1916. ‘Bul. Bur. Stds., 7, p. 252, 1916. ’ Bul. BUY. Stds., 13, p. 355, 1916. 399
W. W.
400
COBLEST~.
[J.F. I.
Inspection of the published curve 4 illustrating these data shows a divergence in the region of 0.48~. Computations made at that time indicated that for the regio’n of 0.48~ to 0.75~ the spectral energy distribution was that of a black body at about 2100' C. (2370’ K.) ; though at that time no significance was attached to it, and hence no mention was made of the fact. The data presented in the present paper are based upon a careful study of the original spectral energy curve. Some of the deviations noted in columns 2 and 3 of Table I resulted from variations ‘in reading the data from the curves. The data at 0.4~ td 0.5~ obtained with the quartz prism are now given, but little weight, owing to the quesbion of aperture correction. The revised data are given in column 3, while column 2 gives the data as previously published. They are practically the same except in the violet. In a recent inter-laboratory comparison, by color matching the acetylene flame against a tungsten lamp, Hyde and his collaborators 5 have placed the color temperature of the Eastman Kodak standard acetylene flame at 2360’ K. Within the experimental errors of observation the spectral energy distribution of several series of observations of this flame was found to be the same as that of the burner (with 8 mm. slit) used by the writer in the visibility of radiation measurements. The spectral energy curve of a black body of 2360” K. was kindly computed by Mr. Herbert Kahler and carefully superposed upon the observed data, using a scale which was sufficiently large to eliminate errors (0.5 per cent.) in drawing the curves and reading off the data. These data are given in column 4. Column 5 gives the ratios of observed to computed data for 2360’ K. It will be noticed that between 0.48~ and 0.75~~ which is the region in which the radiometric measurements are of importance in the visibility of radiation work, there is as close agreement (+2 per cent.) between the computed and observed data as can be expected-for, when the errors incident to color match as well as radiometric observations are considered, it will no doubt be conceded that the color match, as well as the radiometric work, has its limitations. While the writer makes mental reservations as to the efficiency ‘Rul.
BUY. St&.,
“This
JOURNAL,
13, p. 361, 188,
p. Izg,
Fig. 3. 1919.
Sept..1919.1 SPECTRAL
ENERGV IN ACETYLENE FLAME.
-lo1
of the color match test in the extreme violet of the spectrum where the eye is very insensitive, and subject to great decrease in sensitivity with age, he concurs in the recommendation made by Hyde that the spectral energy distribution of the acetylene flame (using a certain type of burner) in the visible spectrum is satisfactorily represented by the black body curve at 2360’ K., as given in Table I. A difference of IO’ or even 50’ K. is hardly TABLE I. Spectral Energy
Transmission of a Cylindrical Acetylene Flame; Observed and Computed Using Wien’s Eq., C=I&~O and T =z.~60” K. E’
Wave
Computed
length
-
--____ 11.5 13.0 16.0 21.9 27.9 29.5 35.0 38.9 42.9 49.8 52.2 62.1 73.0
'450 .460 ,475 ,500
,520
,525
549 :$Z: 575 :ZZ
,620 ,625 ,640 ,650 ,660 ,675 .680 ,700 720
,725 .74o ,750
ii:: 91. I 97.4 107.5 110.9 124.6 138.5 141.9 152.0 158.9
IO. 12. 15.9 21 .o 27.5 29.2 ii:; 42.9 ;Z 62.5 73.3 76.1 85.0 91.2 97.6. 107.5 110.9 124.1 135.7 141 .o 151 .o 157.9
9.2
II.2 14.6 21 .I 27.3 29.2 33::; 43.1 49.9 52.4 62.9
E
-
_
92. 93. 97.3 100.5 99.3
100.0 100.0 e99.8 100.4 100.2 100.3 100.6 101.
$1: 86.0 92. I 98-5 108.0 111.3 124.1 137.2 140.5 150.2 157.2
I
100.8 101. 101
I .o
100.7 100.4 100.7 100.0 99.8 99.6 99.5 99.5
to be considered in view of the great divergence in burners, the effect of humidity, quality of gas, etc. The independent check by Hyde and his collaborators confirms the direct radiometric observations on the spectral energy distribution of the acetylene flame, in the region of 0.5~ to 0.75r, as used in the visibility of radiation work and leaves the visibility data 6 unchanged. WASHINGTON, D. C.,
July
24, IgIg.
“Bul. Bur. Stds., 14, p. 168, 1917.